A recent report by CrowdStrike has found that only 39% of security professionals believe that the security benefits of generative AI outweigh the potential risks. The survey, which included 1,022 security researchers and practitioners from various regions, revealed that while 64% have either purchased or are researching generative AI tools, the majority remain cautious.
The main motivation for adopting generative AI among security professionals is to improve their ability to respond to and defend against cyberattacks. Many are seeking generative AI tools that are partnered with security expertise, rather than general AI solutions. While 40% of respondents believe the rewards and risks of generative AI are comparable, 39% feel that the rewards outweigh the risks.
Measuring ROI when adopting generative AI products has been a challenge for many organizations. CrowdStrike identified quantifying ROI as the top economic concern among respondents, followed by the cost of licensing AI tools and unpredictable pricing models. The report suggests that adding AI to an existing platform rather than purchasing a standalone product could lead to incremental savings.
On the flip side, security professionals are also concerned about securing generative AI itself. Issues such as data exposure to LLMs behind AI products and attacks launched against generative AI tools are top concerns. Nearly all respondents stated that their organizations have implemented new security policies or are in the process of developing policies to govern generative AI within the next year.
Despite these concerns, organizations can leverage generative AI to protect against cyber threats through threat detection and analysis, automated incident response, phishing detection, enhanced security analytics, and synthetic data for training. However, it is crucial for organizations to prioritize safety and privacy controls when purchasing generative AI tools to protect sensitive data, comply with regulations, and mitigate risks. Without proper safeguards, AI tools could expose vulnerabilities and lead to financial, legal, and reputational damage.